Refactored out sanitise, added loading of English letter counts

[cipher-tools.git] / cipher.py

diff --git a/cipher.py b/cipher.py
index f938f3677f321d02805688ac443328314af0406f..b3ab488db74c3af8090aefcc2396c2d503725064 100644 (file)
--- a/cipher.py
+++ b/cipher.py
@@ -1,5 +1,21 @@
 import string
+import collections
 
+english_counts = collections.defaultdict(int)
+with open('count_1l.txt', 'r') as f:
+    for line in f:
+        (letter, count) = line.split("\t")
+        english_counts[letter] = int(count)
+
+def sanitise(text):
+    sanitised = [c.lower() for c in text if c in string.ascii_letters]
+    return ''.join(sanitised)
+
+def letter_frequencies(message):
+    frequencies = collections.defaultdict(int)
+    for letter in sanitise(message): 
+        frequencies[letter]+=1
+    return frequencies
 
 def caesar_cipher_letter(letter, shift):
     if letter in string.ascii_letters: